Abstract The Sahel is one of the most vulnerable regions to climate change. Robust estimation of future changes in the Sahel monsoon is therefore essential for effective climate change adaptation. Unfortunately, state-of-the-art climate models show large uncertainties in their projections of Sahel rainfall. In this study, we use 32 models from CMIP6 to identify the sources of this large intermodel spread of Sahel rainfall. By using maximum covariance analysis, we first highlight two new key drivers of this spread during boreal summer: the interhemispheric temperature gradient and equatorial Pacific sea surface temperature (SST) changes. This contrasts with previous studies, which have focused mainly on the Northern Hemisphere rather than the global scale, and in which the Pacific Ocean has been neglected in favor of the Atlantic. Next, we unravel the physical mechanisms behind these statistical relationships. First, the modulation of the interhemispheric temperature gradient across the models leads to varying latitudinal positions of the intertropical convergence zone and, consequently, varying Sahel rainfall intensity. Second, models that exhibit less warming than the multimodel mean in the equatorial Pacific, thereby projecting a less “El Niño–like” mean state, simulate enhanced precipitation over the central Sahel in the future through modulations of the Walker circulation, the tropical easterly jet, the meridional tropospheric temperature gradient, and hence regional zonal wind shear. Finally, we show that these two indices collectively explain 62% of Sahel rainfall change uncertainty: 40% due to the interhemispheric temperature gradient and 22% through equatorial Pacific SST.